FLUX.1 [schnell]

Fast 12B text-to-image model — high-quality images in 1–4 steps, free for commercial use.

Overview

FLUX.1 [schnell] is the distilled, Apache-licensed member of the FLUX.1 family. It produces strong prompt-following images in as few as four steps. This deployment loads the FP8 all-in-one checkpoint into ComfyUI.

Key features

• ✓
  1–4 step generation
• ✓
  Excellent prompt adherence and text rendering
• ✓
  Apache 2.0 — commercial use OK

Use cases

• →
  Marketing and product imagery
• →
  Concept art
• →
  Rapid prototyping of visuals

On this template

Runs on the ComfyUI template, exposing its ComfyUI at a private HTTPS address. You also get root SSH and JupyterLab.

COMFY_DOWNLOADS

Comfy-Org/flux1-schnell|flux1-schnell-fp8.safetensors|checkpoints

Quick start

1. Click Deploy and pick an available GPU — only offers with enough VRAM are shown.
2. Wait for the status to turn running. The first start downloads the model, which can take a few minutes.
3. Use the connection details on your rental page:

Open https://r12-abcd1234-app.gpu.clearlayer.ca
Gateway login: user  edge   password  <app-token>
Workflow → Browse Templates, pick the template for your model, Queue.